\begin{sidewaystable}[h]
\centering
\caption{Effect of promotion on improved- and LPG-stove adoption in matched NGO/non-NGO villages\label{app:improved_stoves_tab}}
\begin{adjustbox}{max width=\textwidth}
\begin{threeparttable}
\begin{tabular}{lcccccccc}
\toprule
& (1) & (2) & (3) & (4) & (5) & (6) & (7) & (8) \\ \cmidrule(lr){2-3} \cmidrule(lr){4-5} \cmidrule(lr){6-7} \cmidrule(lr){8-9}
& \multicolumn{2}{c}{$\mathbbm{1} \left( \text{Owns improved stove} \right)$} &  \multicolumn{2}{c}{$\mathbbm{1} \left( \text{Uses improved stove} \right)$} & \multicolumn{2}{c}{$\mathbbm{1} \left( \text{Owns LPG stove} \right)$} & \multicolumn{2}{c}{$\mathbbm{1} \left( \text{Uses LPG stove} \right)$} \\  \midrule
\({POST}_1\)&       0.016   &      -0.048   &      0.0039   &      -0.054   &       0.031   &      -0.034   &       0.019   &      -0.041   \\
            &     (0.036)   &     (0.042)   &     (0.034)   &     (0.041)   &     (0.034)   &     (0.037)   &     (0.032)   &     (0.036)   \\ \addlinespace
\({POST}_2\)&        0.14***&        0.12***&        0.12***&       0.095*  &        0.14***&        0.12***&        0.13***&       0.095** \\
            &     (0.039)   &     (0.044)   &     (0.040)   &     (0.048)   &     (0.038)   &     (0.042)   &     (0.038)   &     (0.043)   \\ \addlinespace
\({TREATMENT}_j \times {POST}_1\)&        0.34***&        0.38***&        0.29***&        0.30***&       0.074*  &        0.12***&       0.084** &        0.12***\\
            &     (0.047)   &     (0.066)   &     (0.042)   &     (0.056)   &     (0.038)   &     (0.045)   &     (0.036)   &     (0.044)   \\ \addlinespace
\({TREATMENT}_j \times {POST}_2\)&        0.26***&        0.28***&        0.18***&        0.20***&       0.023   &       0.046   &       0.024   &       0.058   \\
            &     (0.048)   &     (0.065)   &     (0.047)   &     (0.061)   &     (0.042)   &     (0.050)   &     (0.043)   &     (0.050)   \\ \addlinespace
\({NGO}_j \times {POST}_1\)&               &        0.15** &               &        0.14** &               &        0.15** &               &        0.14** \\
            &               &     (0.068)   &               &     (0.063)   &               &     (0.060)   &               &     (0.056)   \\ \addlinespace
\({NGO}_j \times {POST}_2\)&               &       0.040   &               &       0.067   &               &       0.065   &               &       0.076   \\
            &               &     (0.083)   &               &     (0.085)   &               &     (0.082)   &               &     (0.082)   \\ \addlinespace
\({TREATMENT}_j \times {NGO}_j \times {POST}_1\)&               &       -0.10   &               &      -0.045   &               &       -0.12*  &               &      -0.100   \\
            &               &     (0.092)   &               &     (0.080)   &               &     (0.070)   &               &     (0.065)   \\ \addlinespace
\({TREATMENT}_j \times {NGO}_j \times {POST}_2\)&               &      -0.047   &               &      -0.056   &               &      -0.056   &               &      -0.078   \\
            &               &      (0.10)   &               &     (0.098)   &               &     (0.090)   &               &     (0.090)   \\
\midrule
Baseline non-NGO control mean &      0.36 &      0.36 &      0.35 &      0.35 &      0.33 &      0.33 &      0.33 &      0.33  \\
Household fixed-effects & Yes & Yes & Yes & Yes & Yes & Yes & Yes & Yes \\
\(N\)       &       2,829   &       2,829   &       2,829   &       2,829   &       2,829   &       2,829   &       2,829   &       2,829   \\
Adjusted \(R^2\)&        0.56   &        0.57   &        0.52   &        0.53   &        0.65   &        0.65   &        0.64   &        0.64   \\
\bottomrule
\end{tabular}
\begin{tablenotes}
{\setlength\labelsep{0pt}
\footnotesize
\item \textit{Notes}. The outcome variables are ownership (columns 1–2) and use (columns 3–4) of all improved stoves as measured during baseline, and midline and endline follow-up surveys (conducted approximately three and fifteen months after the start of the intervention, respectively). The outcome variable for ownership is an indicator that equals 1 if household \(i\) in hamlet \(j\) reports owning at least one improved stove in survey round \(t\); for use, it is an indicator that equals 1 if the household---conditional on ownership---reports having used such a device in the past week. The even-numbered columns present the results of estimating the triple-differences specification outlined in Equation \eqref{e:triple-diff}. The accompanying odd-numbered columns present results from estimating a double-differences specification without NGO-specific interactions for comparison. 'Improved stove' includes stoves fuelled by biogas, electricity, LPG, kerosene, and commercially available efficient biomass cookstoves; we also include the two ICS promoted as part of the promotion intervention in this definition. Standard errors (in parentheses) are clustered at the hamlet level. \sym{*} \(p<0.10\), \sym{**} \(p<0.05\), \sym{***} \(p<0.01\).}
\end{tablenotes}
\end{threeparttable}
\end{adjustbox}
\end{sidewaystable}
